Study setting

The participants will be recruited from secondary and tertiary referral hospitals in Denmark, Finland and Sweden with at least two practising specialists in hand surgery (online supplemental material I, table S3). To be eligible, investigators must hold a specialist degree in hand surgery or orthopaedics and have performed a minimum of 30 wrist arthroscopies before the trial commencement. One of the hand surgeons will serve as the centre’s co-principal investigator (Co-PI). Each trial centre will be responsible for screening all patients referred under the diagnosis of ulnar wrist pain.

Participants

The Co-PI or deputy investigator (DI) at each centre will screen all patients referred to the trial centre for ulnar wrist pain and record their details in the screening log (figure 2). The inclusion and exclusion criteria have been designed to enable the best possible generalisability of the trial, but exclusion criteria are planned to exclude significant possible sources of invalid data (table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Recruitment, data collection and randomisation.*Based on the referral, an appointment at the outpatient clinic is scheduled for the patient. The cover letter, along with patient information and consent forms, can be sent to the participant along with the details of the outpatient clinic visit.^†If a TFCC rupture is suspected, an MRI can be optionally conducted before the outpatient clinic visit.^‡If suspicion of a TFCC rupture arises during the outpatient clinic visit, the participant will be informed about the trial. If the patient is considered potentially eligible, consent for participation may be obtained. The Co-PI or DI will gather information on grip strength, ROM of the wrist and forearm, and ulnar variance. Additionally, the participant will be given EQ-5D-3L, PRWE, and VAS questionnaires to complete and return before the scheduled arthroscopy.^**Preoperatively, RN, or optionally the Co-PI, or DI, will gather information on sex, age, duration of symptoms, history of smoking, occupation, job, involved hand, hand dominance, and any previous injuries or treatments related to the symptomatic hand. The RN, or optionally the Co-PI or DI, will also ensure that EQ-5D-3L, PRWE, and VAS questionnaires are correctly completed and collect the forms. If a TFCC tear is identified as 1A, 1B, or 1D3 during arthroscopy, and there are no exclusion criteria, the participant's recruitment is considered finalized.^††Randomisation will be conducted during the arthroscopy using the online randomisation software available at https://www.randomizer.at.Co-PI, Co-principal investigators;DI, deputy investigator; EQ-5D-3L, European quality of life 5 dimensions 3 level version; MRI, magnetic resonance image; PRWE, Patient rated wrist evaluation; RN, research nurse; ROM, range of motion; TFCC, triangular fibrocartilage complex; VAS, visual analogue scale.

Patients should present with symptoms and clinical examination findings consistent with a TFCC tear. They must be of working age, willing to participate and able to provide and complete all required information. The TFCC tear, confirmed via arthroscopy, should account for the patient’s symptoms, with ulnocarpal ligament tears being excluded. Patients with remarkable gross instability of the DRUJ, osteoarthritis on the ulnar side of the wrist, inflammatory diseases affecting the wrist or notable ulnar-side instability in structures other than the TFCC will also be excluded.

The aim of the inclusion and exclusion criteria is to identify patients most likely to benefit from surgical treatment of a TFCC tear. Paediatric patients are also excluded because their treatment differs from that of the adult population. Patients with gross instability are excluded because, in the authors opinion of this protocol, they should be evaluated in a separate trial, as the present trial does not have sufficient power to allow for adequate subgroup analysis. Ulnocarpal ligament tears are excluded due to their extreme rarity and the lack of consensus on their treatment.

Patients eligible for the trial must have a suspected TFCC tear based on their medical history and clinical examination, with an optional MRI showing a possible TFCC tear. If they meet the inclusion criteria and none of the exclusion criteria, diagnostic arthroscopy is indicated. Decision of diagnostic arthroscopy will be made by the recruiting hand surgeon.

Interventions and adherence

All eligible patients are provided with a written information form about the trial, and the Co-PI or DI (recruiter) is responsible for obtaining consent from potential participants. The final inclusion or exclusion of participants occurs during an arthroscopic evaluation of the wrist, when a TFCC tear is identified as either (1) central or radial tear or (2) ulnar tear, confirming adherence to the trial’s inclusion criteria. After the diagnostic part of the arthroscopy, the surgeon retrieves the randomisation code from the internet-based programme at https://www.randomizer.at.

In both randomisation cohorts, arthroscopy will be conducted with or without irrigation, and the treatment will be conducted based on the randomisation. For ulnar tears, those undergoing repair will undergo postoperatively 6 weeks of immobilisation with cast, splint or orthosis, which immobilises the forearm and wrist, while those in the non-operative (physiotherapy) group will have a 2-week immobilisation period with dorsal cast, splint or orthosis immobilising the wrist (online supplemental material I, table S5 and online supplemental material II).

The investigated treatments are widely used and demonstrated to be safe.9 10 We see no reason to discontinue or modify the interventions randomly assigned to the participants.

Full adherence to the allocated treatment is anticipated, as arthroscopic debridement and repair of TFCC ulnar tear are performed immediately after randomisation during the same operation. In the central or radial tear randomisation cohort, participants, caregivers and all trial personnel, excluding operating theatre staff, remain blinded to the treatment allocation. To enhance adherence to follow-up, participants receive comprehensive information about the trial and treatments during the initial contact. Participants receive contact details for the coordinating research assistant (CRA) and outpatient clinic at each centre, allowing them to reach out at any point during the trial. Active monitoring of participant controls occurs at specified intervals by junior investigators (JI), research nurses and the CRA. If the participants do not adhere to the follow-up schedule, they are contacted, and patient-reported outcomes are collected via phone if the participant agrees.

Study outcomes

Follow-up time points and collected data in each point are planned so that all the needed data can be gathered while the participants are not overburdened with data collection (table 2).

Sample size

This trial is designed as a superiority trial, aiming to detect a mean difference of 14 points. With type I error rate of 0.05 and a power of 80%, we need 44 participants per arm to detect a difference of >14 points in PRWE assuming SD of 20. Considering attrition rate of 15%, the final number of participants per randomisation cohort arm is 51, totalling 204 participants for the whole trial.

Participant recruitment

The multinational and multicentre setup of this trial is designed to facilitate the recruitment of the required number of participants. CRA will actively monitor the recruitment progress from each centre. If any centre encounters challenges in recruitment, the CRA will promptly contact the JI and PI. Together with the Co-PI, they will engage in discussions to devise strategies for improving recruitment.

The first participant in the trial was recruited on 27 October 2020, and the last is expected to be recruited by the end of 2025. Therefore, the 10-year follow-up period is projected to conclude by the end of 2035. However, if the planned sample size has not been reached by the end of 2025, recruitment will continue until the required number of participants is enrolled, and the end date for the 10-year follow-up will be adjusted accordingly.

Randomisation and blinding

In this trial, a centralised allocation system will be employed. The concealment of allocation is ensured, as the randomisation code will be released only after the diagnosis is confirmed during the arthroscopy. The participants within randomisation cohorts are allocated 1:1 with a random block size. Dominant/non-dominant hand will be used as a stratification criterion.

The operating surgeon and operating theatre staff will not be blinded to the allocation due to the nature of the interventions. However, outcome assessors measuring grip strength and ROMs will be blinded in central or radial tear randomisation cohort and they will not participate in the care at any other point in the trial, except during follow-up visits. For central and radial tears, participants are also blinded, resulting in a triple-blinded randomisation cohort since investigators are also blinded in the data analysis phase as in foveal tear randomisation cohort.

In the central or radial tear cohort, to ensure blinding of the participants, the participants are not able to see the operation area or the monitor. They will listen to music with noise-cancelling headphones throughout the operation. In the placebo group, the operative time will be matched, with the surgeon simulating a debridement procedure. The postoperative protocol is identical in both arms. The staff in the operation theatre will not communicate with the participant after the operation, and they will inform the staff responsible for follow-up in a similar way irrespective of the allocation.

In the ulnar tear cohort, the randomisation between repair and physiotherapy cannot be blinded due to differing postoperative treatments: a 6-week immobilisation in the repair group and a 2-week immobilisation in the physiotherapy group. Maintaining similar postoperative protocols might compromise the distinct outcomes associated with each intervention.

The success of blinding with the participant and the outcome assessor

In the central or radial tear randomisation cohort, during the 1 -year follow-up visit, participants and outcome assessors will be asked to indicate which treatment group they believe the participant belongs to (debridement or placebo surgery). The success of blinding will be reported as a percentage of correct responses in both arms.

Cross-over and unblinding

During recruitment, the Co-PI informs participants that if they do not achieve adequate symptom improvement at the 1-year postoperative evaluation or thereafter, participants in the central or radial tear cohort may be unblinded and undergo debridement if the initial treatment is placebo surgery. In the ulnar tear cohort, the participant may undergo TFCC repair if the initial treatment was physiotherapy. If the initial treatment in the central or radial tear cohort was debridement, or in the ulnar tear cohort it was repaired, the surgeon will determine the appropriate treatment with the participant. Unblinding will be conducted by Co-PI of the centre and CRA. Unblinded participants will continue in the trial but will be marked as ‘unblinded’ in the results.

Baseline assessment

The baseline assessment encompasses typical demographics, duration of symptoms, involved hand, questionnaires, ulnar variance, ROM, and previous injuries or treatments in symptomatic wrist (online supplemental material I, table S7).

Data collection

Trial data will be collected with an electronic data capture platform, REDCap (Vanderbilt University, Nashville, Tennessee, USA). Assessors will undergo training for platform usage, and the REDCap forms include pictorial instructions for various measurements. If needed, individualised training will be organised for specific measurements, such as wrist and forearm ROM with a goniometer and grip strength with Jamar.

Participants have the option to withdraw from the trial at any point. In the event of participant withdrawal, data collection is discontinued, and the data collected up to the point of withdrawal will be used.

Data management

At each centre, data are collected in a password-protected electronic database (REDCap) using the participant’s trial ID number. All data are stored on servers at the coordinating centre. Only investigators and the Data Safety and Monitoring Committee (DSMC) have access to the final dataset, and no contractual agreements limit such access.

During the analysis phase, data will be verified. The magnitude of all results will be reviewed to ensure they fall within a plausible range, and checks will be performed to avoid double data entries.

Statistical plan

Descriptive statistics will be presented as means with standard deviations (SD) for all approximately normally generated continuous variables. Continuous variables that do not follow a normal distribution but show a highly skewed distribution will be presented as median with interquartile ranges (IQR). Categorical outcomes will be presented as numbers with percentages.

For the primary outcome, the statistical tests will be two sided and a p<0.05 will be considered statistically significant. Confidence interval (CI) will be 95% CI and two sided. For the secondary outcomes and time points, we do not intend to adjust the p value for multiple comparisons, and the analyses are considered exploratory. The use of p-values will be toned down when interpreting the results.

The primary analysis will be based on the intention to treat (ITT) principle. Participants allocated to a treatment group (repair or debridement) should be followed up, assessed and analysed as members of that group, regardless of their adherence to the planned course of treatment. A per-protocol (PP) analysis will be conducted as a sensitivity analysis per the actual treatment received by the participants.

The primary comparison in PRWE between groups will be done using a linear mixed model allowing for repeated measures. Group allocation and time points (6 months and 1 year) will be included as fixed effects and participants as random. Baseline score and hand-dominance (in regard to the repaired side) will also be included as fixed covariates. Group×time interaction will also be included in the model. The mean marginal difference at each time point will be interpreted as a treatment effect. Satterwaithes method is used to estimate df for this. 95% CI are reported for each treatment estimate. Due to the repeated measures mixed model analysis, no missing data imputation will be done. The same analytical approach will be used for all continuous secondary outcomes (VAS pain, QoL, ROM and grip strength).

The number of adverse events is a binary secondary outcome variable. A generalised repeated logistic mixed model analysis will be employed, and the difference in the proportion of the outcome events will be reported based on marginal mean effects for each time point. Global rating is an ordinal variable with seven possible categories. Depending on the final distribution of participant ratings, we will employ ordinal logistic regression separately for each time point. If participant ratings show a skewed distribution to higher categories, we will dichotomise the global improvement variable between no change and little better and use a generalised repeated measure mixed logistic model analysis.

All statistical analyses will be made using the latest Rstudio (R core team, Vienna, Austria) with appropriate packages such as lme4, emmeans and margins. The whole statistical plan is presented in attachment of this protocol (statistical analysis plan, SAP).

Analysis

To minimise any bias in interpreting the findings, the trialist, who will be blinded to the treatment allocation, will perform data analysis. Blinded results (groups A and B in the central/radial tear cohort, and groups C and D in the ulnar tear cohort) will be presented to the writing committee, who will collectively reach a consensus on the interpretation of the findings. Once a consensus is reached, the groups will be unblinded.33

Data safety

We will establish a DSMC comprising one statistician, one clinician experienced in clinical trials, and another with expertise in TFCC tears. The DSMC is independent of any competing interest and will be tasked with safeguarding the interests of trial participants. Responsibilities include assessing the safety and efficacy of interventions, reviewing external evidence affecting risk/benefit balance and monitoring overall trial conduct. The DSMC has the authority to evaluate protocol amendments and recommend permission to continue the trial.

Interim analyses and stopping guidelines

Interim analyses outside the planned follow-up points will not be conducted. All the treatments, except placebo surgery, which likely has fewer risks than the treatments of usual care, are used widely for treating TFCC tears and they are considered equally safe.9 10 As there is no expectation of major differences in the incidence of serious adverse events (SAEs) between the trial groups, no priori stopping rules will be applied.

SAEs and safety

All SAEs will be documented and promptly reported to the DSMC. Additionally, annual reports on SAE will be submitted to the Institutional Review Board (IRB) of Pirkanmaa Hospital District. In this trial, SAE is defined as any event leading to participant hospitalisation or death, directly attributable to the treatment. Other conditions linked to the intervention will be considered adverse events.

Centres are instructed to report SAE immediately to the CRA who notifies the JI, PI, Co-PI and DSMC within 48 hours. Participants are informed to contact their centre’s outpatient clinic if they suspect a possible adverse event. Adverse events are also assessed during follow-up visits.

Centres are equipped to handle emergencies, being on-call units of hand surgery, or participants can be directed to a hand surgeon promptly. Adverse events are documented and reported in the results. There is no need for unblinding in the radial or central tear groups, even if the participant is hospitalised, for example, septic arthritis, as no foreign material is left in either treatment to the wrist joint.

Adverse events are treated similarly to any patient, regardless of trial participation. As the investigated treatments are widely used and considered safe, the anticipated risk of significant SAE is low, equivalent to standard treatment. Since the trial serves routine treatment methods, possible treatment injuries will be compensated by the local patient insurance centre.

The investigators hold the responsibility to make the final decision to end the trial based on the DSMC’s recommendation.

Monitoring

The study will undergo monitoring by an external monitor. Initial monitoring of all centres is scheduled for 2023–2024, with a total of three monitoring visits in each centre throughout the trial. The full monitoring plan will be published as a supplementary document to this article.